Image labeling application with a Dash UI.
-
Start the compute and content services in the MLExchange platform. Before moving to the next step, please make sure that the computing API and the content registry are up and running. For more information, please refer to their respective README files.
-
Start splash-ml
-
Start Data Clinic and MLCoach
-
Create a new Python environment and install dependencies:
conda create -n new_env python==3.11
conda activate new_env
pip install .
-
Create a
.env
file using.env.example
as reference. Update this file accordingly. -
Start example app:
python labelmaker.py
The MLExchange File Manager supports data access through:
-
Loading data from file system: You can access image data located at the
data
folder in the main directory. Currently, the supported formats are: PNG, JPG/JPEG, and TIF/TIFF. -
Loading data from Tiled: Alternatively, you can access data through Tiled by providing a
tiled_server_uri
in the frontend of your application and theTILED_KEY
associated with this server as an environment variable.
More information available at File Manager.
Assigning a new label:
- Select all the images to be labeled
- Choose label to be assigned
Removing an assigned label (un-label):
- Select all the images to be unlabeled
- Click the "un-label" button
Choose MLCoach tab on the right sidebar. This options allows users to label images by using a trained MLCoach model and a given probability threshold.
To label images:
- Choose an MLCoach model from the dropdown. The probability of each label will be shown under each image according to the selected model.
- Click on the label-name (e.g. "Label 1") and set a probability threshold.
- Click "Label with Threshold" button.
The images will be automatically labeled based on the threshold. After which, users can manually un-label and re-label following Label Manually procedures.
For further details on the operation of MLCoach, please refer to its documentation.
Choose Data Clinic tab on the right sidebar. This tab allows users to tag similar images under the same label by using a trained Data Clinic model.
Please follow the instructions in the app sidebar. Likewise, users can also manually un-label and re-label following Label Manually procedures afterwards.
For further details on the operation of Data Clinic, please refer to its documentation.
MLExchange Copyright (c) 2024, The Regents of the University of California, through Lawrence Berkeley National Laboratory (subject to receipt of any required approvals from the U.S. Dept. of Energy). All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
(1) Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
(2) Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
(3) Neither the name of the University of California, Lawrence Berkeley National Laboratory, U.S. Dept. of Energy nor the names of its contributors may be used to endorse or promote products derived from this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.